TTNet: Real-time temporal and spatial video analysis of table tennis

• Published in: arXiv.org 2020-04
• Main Authors: Voeikov, Roman, Falaleev, Nikolay, Baikulov, Ruslan
• Format: Article
• Language: English
• Subjects: Data collection; Datasets; Decision making; Machine learning; Masks; Neural networks; Real time; Semantic segmentation; Semantics; Table tennis; Tennis; Tensors
• ISSN: 2331-8422

We present a neural network TTNet aimed at real-time processing of high-resolution table tennis videos, providing both temporal (events spotting) and spatial (ball detection and semantic segmentation) data. This approach gives core information for reasoning score updates by an auto-referee system. We also publish a multi-task dataset OpenTTGames with videos of table tennis games in 120 fps labeled with events, semantic segmentation masks, and ball coordinates for evaluation of multi-task approaches, primarily oriented on spotting of quick events and small objects tracking. TTNet demonstrated 97.0% accuracy in game events spotting along with 2 pixels RMSE in ball detection with 97.5% accuracy on the test part of the presented dataset. The proposed network allows the processing of downscaled full HD videos with inference time below 6 ms per input tensor on a machine with a single consumer-grade GPU. Thus, we are contributing to the development of real-time multi-task deep learning applications and presenting approach, which is potentially capable of substituting manual data collection by sports scouts, providing support for referees' decision-making, and gathering extra information about the game process.